MNIST-Data-PCA-Example

PCA (Principal Component Analysis) is a dimensional reduction technique for high dimensional datasets while preserving most of the information (variance) of the whole dataset. PCA reproject the data to lower dimensions. In this study, we use MNIST dataset with 784 features(dimensions). We try to determine optimum component of the dataset while preserving 99% percent of the information. We made some calculations and draw graphs to obtain best component number. Now let's look at the codes.